Large Language Models (LLMs) have revolutionized artificial intelligence, driving advancements in machine translation, summarization, and conversational agents. However, their increasing integration into critical societal domains has raised concerns about embedded biases, which can perpetuate stereotypes and compromise fairness. These biases stem from various sources, including historical inequalities in training data, linguistic imbalances, and adversarial manipulation. Despite mitigation efforts, recent studies indicate that LLMs remain vulnerable to adversarial attacks designed to elicit biased responses. This work proposes a scalable benchmarking framework to evaluate LLM robustness against adversarial bias elicitation. Our methodology involves (i) systematically probing models with a multi-task approach targeting biases across various sociocultural dimensions, (ii) quantifying robustness through safety scores using an LLM-as-a-Judge approach for automated assessment of model responses, and (iii) employing jailbreak techniques to investigate vulnerabilities in safety mechanisms. Our analysis examines prevalent biases in both small and large state-of-the-art models and their impact on model safety. Additionally, we assess the safety of domain-specific models fine-tuned for critical fields, such as medicine. Finally, we release a curated dataset of bias-related prompts, CLEAR-Bias, to facilitate systematic vulnerability benchmarking. Our findings reveal critical trade-offs between model size and safety, aiding the development of fairer and more robust future language models.
Accurate diagnosis of Mendelian diseases is crucial for precision therapy and assistance in preimplantation genetic diagnosis. However, existing methods often fall short of clinical standards or depend on extensive datasets to build pretrained machine learning models. To address this, we introduce an innovative LLM-Driven multi-agent debate system (MD2GPS) with natural language explanations of the diagnostic results. It utilizes a language model to transform results from data-driven and knowledge-driven agents into natural language, then fostering a debate between these two specialized agents. This system has been tested on 1,185 samples across four independent datasets, enhancing the TOP1 accuracy from 42.9% to 66% on average. Additionally, in a challenging cohort of 72 cases, MD2GPS identified potential pathogenic genes in 12 patients, reducing the diagnostic time by 90%. The methods within each module of this multi-agent debate system are also replaceable, facilitating its adaptation for diagnosing and researching other complex diseases.
We demonstrate how AI agents can coordinate to deceive oversight systems using automated interpretability of neural networks. Using sparse autoencoders (SAEs) as our experimental framework, we show that language models (Llama, DeepSeek R1, and Claude 3.7 Sonnet) can generate deceptive explanations that evade detection. Our agents employ steganographic methods to hide information in seemingly innocent explanations, successfully fooling oversight models while achieving explanation quality comparable to reference labels. We further find that models can scheme to develop deceptive strategies when they believe the detection of harmful features might lead to negative consequences for themselves. All tested LLM agents were capable of deceiving the overseer while achieving high interpretability scores comparable to those of reference labels. We conclude by proposing mitigation strategies, emphasizing the critical need for robust understanding and defenses against deception.
We present a novel, open-source social network simulation framework, MOSAIC, where generative language agents predict user behaviors such as liking, sharing, and flagging content. This simulation combines LLM agents with a directed social graph to analyze emergent deception behaviors and gain a better understanding of how users determine the veracity of online social content. By constructing user representations from diverse fine-grained personas, our system enables multi-agent simulations that model content dissemination and engagement dynamics at scale. Within this framework, we evaluate three different content moderation strategies with simulated misinformation dissemination, and we find that they not only mitigate the spread of non-factual content but also increase user engagement. In addition, we analyze the trajectories of popular content in our simulations, and explore whether simulation agents' articulated reasoning for their social interactions truly aligns with their collective engagement patterns. We open-source our simulation software to encourage further research within AI and social sciences.
Generative AI is transforming computing education by enabling the automatic generation of personalized content and feedback. We investigate its capabilities in providing high-quality programming tasks to students. Despite promising advancements in task generation, a quality gap remains between AI-generated and expert-created tasks. The AI-generated tasks may not align with target programming concepts, could be incomprehensible for students to solve, or may contain critical issues such as incorrect tests. Existing works often require interventions from human teachers for validation. We address these challenges by introducing PyTaskSyn, a novel synthesis technique that first generates a programming task and then decides whether it meets certain quality criteria to be given to students. The key idea is to break this process into multiple stages performed by expert and student agents simulated using both strong and weaker generative models. Through extensive evaluation, we show that PyTaskSyn significantly improves task quality compared to baseline techniques and showcases the importance of each specialized agent type in our validation pipeline. Additionally, we conducted user studies using our publicly available web application and show that PyTaskSyn can deliver high-quality programming tasks comparable to expert-designed ones while reducing workload and costs, and being more engaging than programming tasks that are available in online resources.
Large Language Models (LLMs) are emerging as promising tools for automated reinforcement learning (RL) reward design, owing to their robust capabilities in commonsense reasoning and code generation. By engaging in dialogues with RL agents, LLMs construct a Reward Observation Space (ROS) by selecting relevant environment states and defining their internal operations. However, existing frameworks have not effectively leveraged historical exploration data or manual task descriptions to iteratively evolve this space. In this paper, we propose a novel heuristic framework that enhances LLM-driven reward design by evolving the ROS through a table-based exploration caching mechanism and a text-code reconciliation strategy. Our framework introduces a state execution table, which tracks the historical usage and success rates of environment states, overcoming the Markovian constraint typically found in LLM dialogues and facilitating more effective exploration. Furthermore, we reconcile user-provided task descriptions with expert-defined success criteria using structured prompts, ensuring alignment in reward design objectives. Comprehensive evaluations on benchmark RL tasks demonstrate the effectiveness and stability of the proposed framework. Code and video demos are available at jingjjjjjie.github.io/LLM2Reward.
Whether related to machine learning models' epistemic opacity, algorithmic classification systems' discriminatory automation of testimonial prejudice, the distortion of human beliefs via the 'hallucinations' of generative AI, the inclusion of the global South in global AI governance, the execution of bureaucratic violence via algorithmic systems, or located in the interaction with conversational artificial agents epistemic injustice related to AI is a growing concern. Based on a proposed general taxonomy of epistemic injustice, this paper first sketches a taxonomy of the types of epistemic injustice in the context of AI, relying on the work of scholars from the fields of philosophy of technology, political philosophy and social epistemology. Secondly, an additional perspective on epistemic injustice in the context of AI: generative hermeneutical erasure. I argue that this injustice that can come about through the application of Large Language Models (LLMs) and contend that generative AI, when being deployed outside of its Western space of conception, can have effects of conceptual erasure, particularly in the epistemic domain, followed by forms of conceptual disruption caused by a mismatch between AI system and the interlocutor in terms of conceptual frameworks. AI systems' 'view from nowhere' epistemically inferiorizes non-Western epistemologies and thereby contributes to the erosion of their epistemic particulars, gradually contributing to hermeneutical erasure. This work's relevance lies in proposal of a taxonomy that allows epistemic injustices to be mapped in the AI domain and the proposal of a novel form of AI-related epistemic injustice.
We present Kimi-VL, an efficient open-source Mixture-of-Experts (MoE) vision-language model (VLM) that offers advanced multimodal reasoning, long-context understanding, and strong agent capabilities - all while activating only 2.8B parameters in its language decoder (Kimi-VL-A3B). Kimi-VL demonstrates strong performance across challenging domains: as a general-purpose VLM, Kimi-VL excels in multi-turn agent tasks (e.g., OSWorld), matching flagship models. Furthermore, it exhibits remarkable capabilities across diverse challenging vision language tasks, including college-level image and video comprehension, OCR, mathematical reasoning, and multi-image understanding. In comparative evaluations, it effectively competes with cutting-edge efficient VLMs such as GPT-4o-mini, Qwen2.5-VL-7B, and Gemma-3-12B-IT, while surpassing GPT-4o in several key domains. Kimi-VL also advances in processing long contexts and perceiving clearly. With a 128K extended context window, Kimi-VL can process diverse long inputs, achieving impressive scores of 64.5 on LongVideoBench and 35.1 on MMLongBench-Doc. Its native-resolution vision encoder, MoonViT, further allows it to see and understand ultra-high-resolution visual inputs, achieving 83.2 on InfoVQA and 34.5 on ScreenSpot-Pro, while maintaining lower computational cost for common tasks. Building upon Kimi-VL, we introduce an advanced long-thinking variant: Kimi-VL-Thinking. Developed through long chain-of-thought (CoT) supervised fine-tuning (SFT) and reinforcement learning (RL), this model exhibits strong long-horizon reasoning capabilities. It achieves scores of 61.7 on MMMU, 36.8 on MathVision, and 71.3 on MathVista while maintaining the compact 2.8B activated LLM parameters, setting a new standard for efficient multimodal thinking models. Code and models are publicly accessible at https://github.com/MoonshotAI/Kimi-VL.
Supporting learners' understanding of taught skills in online settings is a longstanding challenge. While exercises and chat-based agents can evaluate understanding in limited contexts, this challenge is magnified when learners seek explanations that delve into procedural knowledge (how things are done) and reasoning (why things happen). We hypothesize that an intelligent agent's ability to understand and explain learners' questions about skills can be significantly enhanced using the TMK (Task-Method-Knowledge) model, a Knowledge-based AI framework. We introduce Ivy, an intelligent agent that leverages an LLM and iterative refinement techniques to generate explanations that embody teleological, causal, and compositional principles. Our initial evaluation demonstrates that this approach goes beyond the typical shallow responses produced by an agent with access to unstructured text, thereby substantially improving the depth and relevance of feedback. This can potentially ensure learners develop a comprehensive understanding of skills crucial for effective problem-solving in online environments.
Multi-agent system (MAS) has demonstrated exceptional capabilities in addressing complex challenges, largely due to the integration of multiple large language models (LLMs). However, the heterogeneity of LLMs, the scalability of quantities of LLMs, and local computational constraints pose significant challenges to hosting these models locally. To address these issues, we propose a new framework termed Distributed Multi-Agent System (DMAS). In DMAS, heterogeneous third-party agents function as service providers managed remotely by a central MAS server and each agent offers its services through API interfaces. However, the distributed nature of DMAS introduces several concerns about trustworthiness. In this paper, we study the Achilles heel of distributed multi-agent systems, identifying four critical trustworthiness challenges: free riding, susceptibility to malicious attacks, communication inefficiencies, and system instability. Extensive experiments across seven frameworks and four datasets reveal significant vulnerabilities of the DMAS. These attack strategies can lead to a performance degradation of up to 80% and attain a 100% success rate in executing free riding and malicious attacks. We envision our work will serve as a useful red-teaming tool for evaluating future multi-agent systems and spark further research on trustworthiness challenges in distributed multi-agent systems.
We propose a novel framework for generating causal graphs from narrative texts, bridging high-level causality and detailed event-specific relationships. Our method first extracts concise, agent-centered vertices using large language model (LLM)-based summarization. We introduce an "Expert Index," comprising seven linguistically informed features, integrated into a Situation-Task-Action-Consequence (STAC) classification model. This hybrid system, combining RoBERTa embeddings with the Expert Index, achieves superior precision in causal link identification compared to pure LLM-based approaches. Finally, a structured five-iteration prompting process refines and constructs connected causal graphs. Experiments on 100 narrative chapters and short stories demonstrate that our approach consistently outperforms GPT-4o and Claude 3.5 in causal graph quality, while maintaining readability. The open-source tool provides an interpretable, efficient solution for capturing nuanced causal chains in narratives.
Deep reinforcement learning (DRL) has effectively enhanced gameplay experiences and game design across various game genres. However, few studies on fighting game agents have focused explicitly on enhancing player enjoyment, a critical factor for both developers and players. To address this gap and establish a practical baseline for designing enjoyability-focused agents, we propose a two-tier agent (TTA) system and conducted experiments in the classic fighting game Street Fighter II. The first tier of TTA employs a task-oriented network architecture, modularized reward functions, and hybrid training to produce diverse and skilled DRL agents. In the second tier of TTA, a Large Language Model Hyper-Agent, leveraging players' playing data and feedback, dynamically selects suitable DRL opponents. In addition, we investigate and model several key factors that affect the enjoyability of the opponent. The experiments demonstrate improvements from 64. 36% to 156. 36% in the execution of advanced skills over baseline methods. The trained agents also exhibit distinct game-playing styles. Additionally, we conducted a small-scale user study, and the overall enjoyment in the player's feedback validates the effectiveness of our TTA system.
We introduce AgentAda, the first LLM-powered analytics agent that can learn and use new analytics skills to extract more specialized insights. Unlike existing methods that require users to manually decide which data analytics method to apply, AgentAda automatically identifies the skill needed from a library of analytical skills to perform the analysis. This also allows AgentAda to use skills that existing LLMs cannot perform out of the box. The library covers a range of methods, including clustering, predictive modeling, and NLP techniques like BERT, which allow AgentAda to handle complex analytics tasks based on what the user needs. AgentAda's dataset-to-insight extraction strategy consists of three key steps: (I) a question generator to generate queries relevant to the user's goal and persona, (II) a hybrid Retrieval-Augmented Generation (RAG)-based skill matcher to choose the best data analytics skill from the skill library, and (III) a code generator that produces executable code based on the retrieved skill's documentation to extract key patterns. We also introduce KaggleBench, a benchmark of curated notebooks across diverse domains, to evaluate AgentAda's performance. We conducted a human evaluation demonstrating that AgentAda provides more insightful analytics than existing tools, with 48.78% of evaluators preferring its analyses, compared to 27.67% for the unskilled agent. We also propose a novel LLM-as-a-judge approach that we show is aligned with human evaluation as a way to automate insight quality evaluation at larger scale.
Quantum feature maps are a key component of quantum machine learning, encoding classical data into quantum states to exploit the expressive power of high-dimensional Hilbert spaces. Despite their theoretical promise, designing quantum feature maps that offer practical advantages over classical methods remains an open challenge. In this work, we propose an agentic system that autonomously generates, evaluates, and refines quantum feature maps using large language models. The system consists of five component: Generation, Storage, Validation, Evaluation, and Review. Using these components, it iteratively improves quantum feature maps. Experiments on the MNIST dataset show that it can successfully discover and refine feature maps without human intervention. The best feature map generated outperforms existing quantum baselines and achieves competitive accuracy compared to classical kernels across MNIST, Fashion-MNIST, and CIFAR-10. Our approach provides a framework for exploring dataset-adaptive quantum features and highlights the potential of LLM-driven automation in quantum algorithm design.
As Large Language Models (LLMs) become increasingly integrated into real-world, autonomous applications, relying on static, pre-annotated references for evaluation poses significant challenges in cost, scalability, and completeness. We propose Tool-Augmented LLM Evaluation (TALE), a framework to assess LLM outputs without predetermined ground-truth answers. Unlike conventional metrics that compare to fixed references or depend solely on LLM-as-a-judge knowledge, TALE employs an agent with tool-access capabilities that actively retrieves and synthesizes external evidence. It iteratively generates web queries, collects information, summarizes findings, and refines subsequent searches through reflection. By shifting away from static references, TALE aligns with free-form question-answering tasks common in real-world scenarios. Experimental results on multiple free-form QA benchmarks show that TALE not only outperforms standard reference-based metrics for measuring response accuracy but also achieves substantial to near-perfect agreement with human evaluations. TALE enhances the reliability of LLM evaluations in real-world, dynamic scenarios without relying on static references.
As demand for Large Language Models (LLMs) and AI agents rapidly grows, optimizing systems for efficient LLM inference becomes critical. While significant efforts have targeted system-level engineering, little is explored through a mathematical modeling and queuing perspective. In this paper, we aim to develop the queuing fundamentals for LLM inference, bridging the gap between queuing and LLM system communities. In particular, we study the throughput aspect in LLM inference systems. We prove that a large class of 'work-conserving' scheduling algorithms can achieve maximum throughput for both individual requests and AI agent workloads, highlighting 'work-conserving' as a key design principle in practice. Evaluations of real-world systems show that Orca and Sarathi-serve are throughput-optimal, reassuring practitioners, while FastTransformer and vanilla vLLM are not maximally stable and should be used with caution. Our results highlight the substantial benefits queuing community can offer in improving LLM inference systems and call for more interdisciplinary developments.
Large Language Models (LLMs) are increasingly utilized in multi-agent systems (MAS) to enhance collaborative problem-solving and interactive reasoning. Recent advancements have enabled LLMs to function as autonomous agents capable of understanding complex interactions across multiple topics. However, deploying LLMs in MAS introduces challenges related to context management, response consistency, and scalability, especially when agents must operate under memory limitations and handle noisy inputs. While prior research has explored optimizing context sharing and response latency in LLM-driven MAS, these efforts often focus on either fully centralized or decentralized configurations, each with distinct trade-offs. In this paper, we develop a probabilistic framework to analyze the impact of shared versus separate context configurations on response consistency and response times in LLM-based MAS. We introduce the Response Consistency Index (RCI) as a metric to evaluate the effects of context limitations, noise, and inter-agent dependencies on system performance. Our approach differs from existing research by focusing on the interplay between memory constraints and noise management, providing insights into optimizing scalability and response times in environments with interdependent topics. Through this analysis, we offer a comprehensive understanding of how different configurations impact the efficiency of LLM-driven multi-agent systems, thereby guiding the design of more robust architectures.
Reinforcement learning (RL) agents have shown remarkable performances in various environments, where they can discover effective policies directly from sensory inputs. However, these agents often exploit spurious correlations in the training data, resulting in brittle behaviours that fail to generalize to new or slightly modified environments. To address this, we introduce the Causal Object-centric Model Extraction Tool (COMET), a novel algorithm designed to learn the exact interpretable causal world models (CWMs). COMET first extracts object-centric state descriptions from observations and identifies the environment's internal states related to the depicted objects' properties. Using symbolic regression, it models object-centric transitions and derives causal relationships governing object dynamics. COMET further incorporates large language models (LLMs) for semantic inference, annotating causal variables to enhance interpretability. By leveraging these capabilities, COMET constructs CWMs that align with the true causal structure of the environment, enabling agents to focus on task-relevant features. The extracted CWMs mitigate the danger of shortcuts, permitting the development of RL systems capable of better planning and decision-making across dynamic scenarios. Our results, validated in Atari environments such as Pong and Freeway, demonstrate the accuracy and robustness of COMET, highlighting its potential to bridge the gap between object-centric reasoning and causal inference in reinforcement learning.
Agents powered by Large Language Models (LLMs) have recently demonstrated impressive capabilities in various tasks. Still, they face limitations in tasks requiring specific, structured knowledge, flexibility, or accountable decision-making. While agents are capable of perceiving their environments, forming inferences, planning, and executing actions towards goals, they often face issues such as hallucinations and lack of contextual memory across interactions. This paper explores how Case-Based Reasoning (CBR), a strategy that solves new problems by referencing past experiences, can be integrated into LLM agent frameworks. This integration allows LLMs to leverage explicit knowledge, enhancing their effectiveness. We systematically review the theoretical foundations of these enhanced agents, identify critical framework components, and formulate a mathematical model for the CBR processes of case retrieval, adaptation, and learning. We also evaluate CBR-enhanced agents against other methods like Chain-of-Thought reasoning and standard Retrieval-Augmented Generation, analyzing their relative strengths. Moreover, we explore how leveraging CBR's cognitive dimensions (including self-reflection, introspection, and curiosity) via goal-driven autonomy mechanisms can further enhance the LLM agent capabilities. Contributing to the ongoing research on neuro-symbolic hybrid systems, this work posits CBR as a viable technique for enhancing the reasoning skills and cognitive aspects of autonomous LLM agents.
The integration of tool learning with Large Language Models (LLMs) has expanded their capabilities in handling complex tasks by leveraging external tools. However, existing benchmarks for tool learning inadequately address critical real-world personalized scenarios, particularly those requiring multi-hop reasoning and inductive knowledge adaptation in dynamic environments. To bridge this gap, we introduce FamilyTool, a novel benchmark grounded in a family-based knowledge graph (KG) that simulates personalized, multi-hop tool use scenarios. FamilyTool challenges LLMs with queries spanning 1 to 3 relational hops (e.g., inferring familial connections and preferences) and incorporates an inductive KG setting where models must adapt to unseen user preferences and relationships without re-training, a common limitation in prior approaches that compromises generalization. We further propose KGETool: a simple KG-augmented evaluation pipeline to systematically assess LLMs' tool use ability in these settings. Experiments reveal significant performance gaps in state-of-the-art LLMs, with accuracy dropping sharply as hop complexity increases and inductive scenarios exposing severe generalization deficits. These findings underscore the limitations of current LLMs in handling personalized, evolving real-world contexts and highlight the urgent need for advancements in tool-learning frameworks. FamilyTool serves as a critical resource for evaluating and advancing LLM agents' reasoning, adaptability, and scalability in complex, dynamic environments. Code and dataset are available at Github.
Distributed databases are critical infrastructures for today's large-scale software systems, making effective failure management essential to ensure software availability. However, existing approaches often overlook the role distinctions within distributed databases and rely on small-scale models with limited generalization capabilities. In this paper, we conduct a preliminary empirical study to emphasize the unique significance of different roles. Building on this insight, we propose AgentFM, a role-aware failure management framework for distributed databases powered by LLM-driven multi-agents. AgentFM addresses failure management by considering system roles, data roles, and task roles, with a meta-agent orchestrating these components. Preliminary evaluations using Apache IoTDB demonstrate the effectiveness of AgentFM and open new directions for further research.
Large language models (LLMs) offer a promising pre-screening tool, improving early disease detection and providing enhanced healthcare access for underprivileged communities. The early diagnosis of various diseases continues to be a significant challenge in healthcare, primarily due to the nonspecific nature of early symptoms, the shortage of expert medical practitioners, and the need for prolonged clinical evaluations, all of which can delay treatment and adversely affect patient outcomes. With impressive accuracy in prediction across a range of diseases, LLMs have the potential to revolutionize clinical pre-screening and decision-making for various medical conditions. In this work, we study the diagnostic capability of LLMs for Rheumatoid Arthritis (RA) with real world patients data. Patient data was collected alongside diagnoses from medical experts, and the performance of LLMs was evaluated in comparison to expert diagnoses for RA disease prediction. We notice an interesting pattern in disease diagnosis and find an unexpected \textit{misalignment between prediction and explanation}. We conduct a series of multi-round analyses using different LLM agents. The best-performing model accurately predicts rheumatoid arthritis (RA) diseases approximately 95\% of the time. However, when medical experts evaluated the reasoning generated by the model, they found that nearly 68\% of the reasoning was incorrect. This study highlights a clear misalignment between LLMs high prediction accuracy and its flawed reasoning, raising important questions about relying on LLM explanations in clinical settings. \textbf{LLMs provide incorrect reasoning to arrive at the correct answer for RA disease diagnosis.}
Processing structured tabular data, particularly lengthy tables, constitutes a fundamental yet challenging task for large language models (LLMs). However, existing long-context benchmarks primarily focus on unstructured text, neglecting the challenges of long and complex structured tables. To address this gap, we introduce NeedleInATable (NIAT), a novel task that treats each table cell as a "needle" and requires the model to extract the target cell under different queries. Evaluation results of mainstream LLMs on this benchmark show they lack robust long-table comprehension, often relying on superficial correlations or shortcuts for complex table understanding tasks, revealing significant limitations in processing intricate tabular data. To this end, we propose a data synthesis method to enhance models' long-table comprehension capabilities. Experimental results show that our synthesized training data significantly enhances LLMs' performance on the NIAT task, outperforming both long-context LLMs and long-table agent methods. This work advances the evaluation of LLMs' genuine long-structured table comprehension capabilities and paves the way for progress in long-context and table understanding applications.
Building precise simulations of the real world and invoking numerical solvers to answer quantitative problems is an essential requirement in engineering and science. We present FEABench, a benchmark to evaluate the ability of large language models (LLMs) and LLM agents to simulate and solve physics, mathematics and engineering problems using finite element analysis (FEA). We introduce a comprehensive evaluation scheme to investigate the ability of LLMs to solve these problems end-to-end by reasoning over natural language problem descriptions and operating COMSOL Multiphysics$^\circledR$, an FEA software, to compute the answers. We additionally design a language model agent equipped with the ability to interact with the software through its Application Programming Interface (API), examine its outputs and use tools to improve its solutions over multiple iterations. Our best performing strategy generates executable API calls 88% of the time. LLMs that can successfully interact with and operate FEA software to solve problems such as those in our benchmark would push the frontiers of automation in engineering. Acquiring this capability would augment LLMs' reasoning skills with the precision of numerical solvers and advance the development of autonomous systems that can tackle complex problems in the real world. The code is available at https://github.com/google/feabench
Therapeutic development is a costly and high-risk endeavor that is often plagued by high failure rates. To address this, we introduce TxGemma, a suite of efficient, generalist large language models (LLMs) capable of therapeutic property prediction as well as interactive reasoning and explainability. Unlike task-specific models, TxGemma synthesizes information from diverse sources, enabling broad application across the therapeutic development pipeline. The suite includes 2B, 9B, and 27B parameter models, fine-tuned from Gemma-2 on a comprehensive dataset of small molecules, proteins, nucleic acids, diseases, and cell lines. Across 66 therapeutic development tasks, TxGemma achieved superior or comparable performance to the state-of-the-art generalist model on 64 (superior on 45), and against state-of-the-art specialist models on 50 (superior on 26). Fine-tuning TxGemma models on therapeutic downstream tasks, such as clinical trial adverse event prediction, requires less training data than fine-tuning base LLMs, making TxGemma suitable for data-limited applications. Beyond these predictive capabilities, TxGemma features conversational models that bridge the gap between general LLMs and specialized property predictors. These allow scientists to interact in natural language, provide mechanistic reasoning for predictions based on molecular structure, and engage in scientific discussions. Building on this, we further introduce Agentic-Tx, a generalist therapeutic agentic system powered by Gemini 2.5 that reasons, acts, manages diverse workflows, and acquires external domain knowledge. Agentic-Tx surpasses prior leading models on the Humanity's Last Exam benchmark (Chemistry & Biology) with 52.3% relative improvement over o3-mini (high) and 26.7% over o3-mini (high) on GPQA (Chemistry) and excels with improvements of 6.3% (ChemBench-Preference) and 2.4% (ChemBench-Mini) over o3-mini (high).
By 2028 most cybersecurity actions will be autonomous, with humans teleoperating. We present the first classification of autonomy levels in cybersecurity and introduce Cybersecurity AI (CAI), an open-source framework that democratizes advanced security testing through specialized AI agents. Through rigorous empirical evaluation, we demonstrate that CAI consistently outperforms state-of-the-art results in CTF benchmarks, solving challenges across diverse categories with significantly greater efficiency -up to 3,600x faster than humans in specific tasks and averaging 11x faster overall. CAI achieved first place among AI teams and secured a top-20 position worldwide in the "AI vs Human" CTF live Challenge, earning a monetary reward of $750. Based on our results, we argue against LLM-vendor claims about limited security capabilities. Beyond cybersecurity competitions, CAI demonstrates real-world effectiveness, reaching top-30 in Spain and top-500 worldwide on Hack The Box within a week, while dramatically reducing security testing costs by an average of 156x. Our framework transcends theoretical benchmarks by enabling non-professionals to discover significant security bugs (CVSS 4.3-7.5) at rates comparable to experts during bug bounty exercises. By combining modular agent design with seamless tool integration and human oversight (HITL), CAI addresses critical market gaps, offering organizations of all sizes access to AI-powered bug bounty security testing previously available only to well-resourced firms -thereby challenging the oligopolistic ecosystem currently dominated by major bug bounty platforms.
The increasing deployment of intelligent agents in digital ecosystems, such as social media platforms, has raised significant concerns about traceability and accountability, particularly in cybersecurity and digital content protection. Traditional large language model (LLM) watermarking techniques, which rely on token-level manipulations, are ill-suited for agents due to the challenges of behavior tokenization and information loss during behavior-to-action translation. To address these issues, we propose Agent Guide, a novel behavioral watermarking framework that embeds watermarks by guiding the agent's high-level decisions (behavior) through probability biases, while preserving the naturalness of specific executions (action). Our approach decouples agent behavior into two levels, behavior (e.g., choosing to bookmark) and action (e.g., bookmarking with specific tags), and applies watermark-guided biases to the behavior probability distribution. We employ a z-statistic-based statistical analysis to detect the watermark, ensuring reliable extraction over multiple rounds. Experiments in a social media scenario with diverse agent profiles demonstrate that Agent Guide achieves effective watermark detection with a low false positive rate. Our framework provides a practical and robust solution for agent watermarking, with applications in identifying malicious agents and protecting proprietary agent systems.
Large language model-based agents are becoming increasingly popular as a low-cost mechanism to provide personalized, conversational advice, and have demonstrated impressive capabilities in relatively simple scenarios, such as movie recommendations. But how do these agents perform in complex high-stakes domains, where domain expertise is essential and mistakes carry substantial risk? This paper investigates the effectiveness of LLM-advisors in the finance domain, focusing on three distinct challenges: (1) eliciting user preferences when users themselves may be unsure of their needs, (2) providing personalized guidance for diverse investment preferences, and (3) leveraging advisor personality to build relationships and foster trust. Via a lab-based user study with 64 participants, we show that LLM-advisors often match human advisor performance when eliciting preferences, although they can struggle to resolve conflicting user needs. When providing personalized advice, the LLM was able to positively influence user behavior, but demonstrated clear failure modes. Our results show that accurate preference elicitation is key, otherwise, the LLM-advisor has little impact, or can even direct the investor toward unsuitable assets. More worryingly, users appear insensitive to the quality of advice being given, or worse these can have an inverse relationship. Indeed, users reported a preference for and increased satisfaction as well as emotional trust with LLMs adopting an extroverted persona, even though those agents provided worse advice.
We explore the use of Large Language Models (LLMs) for automated assessment of open-text student reflections and prediction of academic performance. Traditional methods for evaluating reflections are time-consuming and may not scale effectively in educational settings. In this work, we employ LLMs to transform student reflections into quantitative scores using two assessment strategies (single-agent and multi-agent) and two prompting techniques (zero-shot and few-shot). Our experiments, conducted on a dataset of 5,278 reflections from 377 students over three academic terms, demonstrate that the single-agent with few-shot strategy achieves the highest match rate with human evaluations. Furthermore, models utilizing LLM-assessed reflection scores outperform baselines in both at-risk student identification and grade prediction tasks. These findings suggest that LLMs can effectively automate reflection assessment, reduce educators' workload, and enable timely support for students who may need additional assistance. Our work emphasizes the potential of integrating advanced generative AI technologies into educational practices to enhance student engagement and academic success.
Enforcing archival standards requires specialized expertise, and manually creating metadata descriptions for archival materials is a tedious and error-prone task. This work aims at exploring the potential of agentic AI and large language models (LLMs) in addressing the challenges of implementing a standardized archival description process. To this end, we introduce an agentic AI-driven system for automated generation of high-quality metadata descriptions of archival materials. We develop a federated optimization approach that unites the intelligence of multiple LLMs to construct optimal archival metadata. We also suggest methods to overcome the challenges associated with using LLMs for consistent metadata generation. To evaluate the feasibility and effectiveness of our techniques, we conducted extensive experiments using a real-world dataset of archival materials, which covers a variety of document types and data formats. The evaluation results demonstrate the feasibility of our techniques and highlight the superior performance of the federated optimization approach compared to single-model solutions in metadata quality and reliability.